1. Survey of Misannotations and
Pseudogenes in the Arabidopsis Genome
Tanmay Prakash

2. Objectives Objectives Find Possible Misannotations
Find Possible Pseudogenes
Why
Misannotation can hinder research
Pseudogenes can be used to study natural selection

3. Misannotations CDS Intron UTR
Many misannotations are the result of gene prediction programs mislabeling introns because of the presence of a stop codon

4. Pseudogenes
Pseudogenes are DNA sequences that no longer function but resemble the functional genes they once were. There are two types:
Processed
Non-processed
Common Properties of Pseudogenes
Stop Codons
Frameshift mutations
Lack of Selective Pressure
Processed:formed by retrotransposition and comprise most of the pseudogenes in mammals
Non-processed:products of duplication of the entirety of portion of a segment of genes followed by mutations. Because polyploidiszation (the process of having more one sets of chromosomes) is common in plants, the majority of pseudogenes in plants are non-processed
Lack of Selective Pressure: Measured using Ka/Ks. Ka(nonsyn) Ks(syn). Functional genes have more syn so Ka/Ks significantly less than one. Pseudogenes don’t care so Ka/Ks significantly closer one.
Because pseudogenes have these stop codons and frameshift mutations, the gene prediction programs often misannotate them
agtacatgcataggactcgatcgactc
STCIGLDRL
agtacatgataggactcgatcgactc
ST..DSID

5. Pipeline Query Protein Domains Genes BLAST Matching Search In Introns
Subject
Arabidopsis
Introns
BLAST
Search
HMMER
CDS
Genes
Matching
In Introns
In CDS
In Both
Possibly
Misannotated
Check for
Stop Codons
Frameshift
Check
Ka/Ks
Possible
Pseudogenes

6. Query Protein Domains Genes BLAST Matching Search In Introns Subject
Arabidopsis
Introns
Query
Protein
Domains
HMMER
Search
Genes
Matching
In Exons
Subject
Arabidopsis
CDS
Each of the 8296 protein domain families is searched against the introns of the genes of the Arabidopsis genome. This finds any introns where there are matches to a protein domain.
This is done also for the coding sequence of the Arabidopsis genome, but using a HMMER search. A HMMER search would’ve been used for the intron search, but it would take far too long. This search found matches to any domains in the coding sequence.

7. Genes Possibly Matching Misannotated In Both Genes
Genes that don’t have matches to the same domain in both the introns and the coding sequence are then filtered out. These genes are possibly misannotated. These genes were further filtered to leave the genes that had matches in an intron and its flanking exons.
These introns will be checked for stop codons and frameshift mutations. The Ka/Ks value will also be checked. This information will be used to identify pseudogenes.

8. Results
There were 346 genes (different models not included) that had matches to the same domain in the introns and exons
There were 299 genes (different models not included) that had matches to the same domain in an intron and flanking exons. These are most likely misannotations.

9. 4 domains with the most possible misannotations

13. Future Research
Identify pseudogenes by looking for stop codons, and frameshift mutations in the introns and checking the Ka/Ks value
Use a more recent database of domains
Follow the same process for the rice genome

14. Acknowledgement Dr. Shin-Han Shiu Dr. Kosuke Hanada
Dr. Melissa Lehti-Shiu
Dr. Gail Richmond
HSHSP